objective-create

objective-create is a command for Claude Code from dagster-io/dagster. It costs 7 tokens per session (2,471 once invoked), scanned A, original, Apache-2.0.

A guided command that turns your description of a goal into a structured objective for you to review and approve.

In plain words
What is it for?
Use it to define coding tasks, explore related parts of a codebase, and prepare work that can be split into implementation phases.
Why use it?
It helps turn scattered notes, constraints, and design decisions into a clear plan with a realistic scope.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/dagster-io/dagster/objective-create
Clone the repo
git clone --depth 1 https://github.com/dagster-io/dagster

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for objective-create

README.md
[![agentmods](https://agentmods.dev/badge/commands/dagster-io/dagster/objective-create.svg)](https://agentmods.dev/commands/dagster-io/dagster/objective-create)
Your own site
<a href="https://agentmods.dev/commands/dagster-io/dagster/objective-create"><img src="https://agentmods.dev/badge/commands/dagster-io/dagster/objective-create.svg" alt="Measured on agentmods" height="20"></a>
Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,471 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00007 $0.02471
Opus 5 $0.00003 $0.01236
Sonnet 5 $0.00001 $0.00494
Haiku 4.5 $0.00001 $0.00247

Measured 3d ago against content hash 27d263451fdf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

objective-create scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/erk/objective-create.md · 371 lines

How it starts

The opening of the file, as written. The whole thing — 371 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/erk:objective-create

Create a new objective through an interactive, guided process. You describe what you want to accomplish, and Claude proposes a structured objective for your approval.

Usage

/erk:objective-create

Agent Instructions

Step 1: Prompt for Description

Ask the user to describe what they want to accomplish:

What do you want to accomplish? Describe it however makes sense to you - goals,
constraints, design decisions you've already made, context about the codebase, etc.

I'll structure it into a formal objective for your review.

Wait for the user's response. They may provide:

  • A brief summary
  • A detailed brain dump
  • Existing notes or documentation
  • A combination of the above

Step 2: Analyze and Explore

Based on the user's description:

  1. Identify key elements:

    • What is the transformation/goal?
    • What design decisions are mentioned or implied?
    • What constraints exist?
    • What's the scope?
  2. Explore the codebase if needed:

    • If the user references specific code, read it to understand context
    • If architecture decisions are mentioned, verify current state
    • Gather enough context to propose realistic phases
  3. Capture context for downstream implementers:

    During objective creation, you gather valuable context that would otherwise be lost: web searches, codebase exploration, user feedback, and your reasoning. This context belongs in the objective issue.

    Ask yourself: "What would a future agent (or human) implementing this need to know?"

    Examples of valuable context:

    • Inventories: Grep results showing all instances of a pattern, categorized by type
    • Root cause analysis: Why the current state exists, what caused the problem
    • Codebase findings: Relevant files, existing patterns, architectural constraints
    • External research: API docs, library behavior, migration guides discovered via web search
    • Design rationale: Why certain approaches were chosen or rejected
    • User preferences: Feedback and decisions made during the conversation

Read the full file on GitHub · 371 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 371 lines · 7 tokens per session scan A 27d263451fdf

Subscribe to this mod's changes

objective-create is a command published in the GitHub repository dagster-io/dagster (16,073 stars, last pushed 5d ago), licensed Apache-2.0. It adds 7 tokens to every session and 2,471 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.